Translesion Synthesis or Repair by Specialized DNA Polymerases Limits Excessive Genomic Instability upon Replication Stress

Abstract

DNA can experience "replication stress", an important source of genome instability, induced by various external or endogenous impediments that slow down or stall DNA synthesis. While genome instability is largely documented to favor both tumor formation and heterogeneity, as well as drug resistance, conversely, excessive instability appears to suppress tumorigenesis and is associated with improved prognosis. These findings support the view that karyotypic diversity, necessary to adapt to selective pressures, may be limited in tumors so as to reduce the risk of excessive instability. This review aims to highlight the contribution of specialized DNA polymerases in limiting extreme genetic instability by allowing DNA replication to occur even in the presence of DNA damage, to either avoid broken forks or favor their repair after collapse. These mechanisms and their key regulators Rad18 and Polθ not only offer diversity and evolutionary advantage by increasing mutagenic events, but also provide cancer cells with a way to escape anti-cancer therapies that target replication forks.

Keywords: DSB repair; POLQ; Pol theta; TMEJ; genome instability; replicative stress; specialized DNA polymerases; translesion synthesis (TLS), Rad18.

Figure 1

Speculative model of translesion synthesis (TLS) pol recruitment at replication forks stalled by DNA distorting lesions. (a) When the replisome encounters DNA distorting lesions (red triangles), replicative polymerases (Pol α, δ, ε) are stalled, while the replicative helicase (CMG complex) can slide through. This process generates an excess of ssDNA (b). Discontinuous DNA synthesis continues on the lagging strand, which generates small replication intermediates that prime the recruitment of specific checkpoint factors leading to activation of the ATR-dependent checkpoint (DDR, not shown). TLS Polκ is recruited onto these replication intermediates to stabilize them and facilitate ATR signaling through Chk1 phosphorylation [38]. DDR activation activates the CRL4^Cdt2 ubiquitin ligase, which induces PCNA-dependent PIP degron-containing proteins destruction, such as the p12 subunit of the Polδ holoenzyme. This process facilitates Polζ4 complex recruitment into the Polδ holoenzyme to position it for bypass extension. Recruitment of RPA onto ssDNA brings about Primpol and other factors (see Figure 2). Primpol allows repriming on the leading strand. Rad18 binding is also stabilized by interaction with Polη and DDK. The Rad6/Rad18 complex then catalyzes PCNA^mUb, thus stabilizing interaction of Polη (and/or other Y-family TLS pols) with PCNA. Polη poly-SUMO modification by PIAS1 and PCNA ISGYlation by the ISG15 (E3) ligase EPF also occurs at this stage. Following this step, the lesion is b-bypassed by the sequential action of Polη and Polζ (insertion and extension, respectively). Upon lesion bypass, Polη is polyubiquitinated by CRL4^Cdt2, OGTylated, and extracted from chromatin by the p97/VPC/Segregase complex to be targeted for proteasomal degradation, while PCNA is deubiquitinated by the concerted action of USP1 and USP10, thus leading to disassembly of the TLS complex including Rad18 (c). The Ubr5 ubiquitin ligase is also implicated in this step by mediating Polη chromatin binding through the ubiquitination of histone H2B [39]. Finally, inactivation of CRL4^Cdt2 and stabilization of the ubiquitin hydrolase USP1 resets PCNA, to a DNA replication-competent form followed by TLS Pol degradation and/or release, thus allowing fork restart.

Figure 2

Factors mediating Rad18 recruitment. (a) Recruitment of Rad52, NBS1 and the RFWD3 ubiquitin ligase occurs onto ssDNA via replication protein A (RPA). Rad52 and NBS1 may bridge the interaction between RPA and Rad18, while RFWD3 might non-specifically polyubiquitinate proteins bound to ssDNA (question mark), thus facilitating TLS factor nucleation. (b) The SART3 pre-mRNA splicing factor may facilitate chromatin remodeling and PCNA^mUb through interaction with Rad18 and Polη when replication forks encounter transcription units (RNA pol2). RNA–DNA hybrids with displaced ssDNA (R-loops) may also be generated in this situation, and thus promote RPA and Rad18 recruitment. (c) DNA–protein crosslinks (DPC) stimulate recruitment of the Spartan metalloprotease through interaction with PCNA. Spartan recruitment will facilitate Rad18 recruitment and PCNA^mUb.

Figure 3

Speculative model of theta-mediated end joining (TMEJ)/Polθ activity as back-up repair of DSB in homologous recombination (HR) deficient cells. DSBs occurring at stalled/collapsed replication forks in S/G2 phases of the cell cycle can be resected leading to short segments of ssDNA which are quickly coated by RPA and can be exchanged for RAD51 when long end-resection occurs, then promoting strand invasion and copying from the sister chromatid in a proficient BRCA-dependent HR pathway. When HR is defective (BRCA mutated genes), or when alternative repair is preferred (choice poorly understood), TMEJ can act as a back-up repair pathway, involving the critical role of Polθ. Polθ-helicase can displace either RPA or RAD51, and Polθ polymerase promotes the synapsis of the opposing ends and performs a bidirectional scanning initiated from the 3′ termini to identify internal microhomologies which can be annealed, thus generating 3′ flaps. Polθ and/or FEN1 can remove the 3′ flaps, and Polθ can start the repair DNA synthesis with poor processivity and frequent aborted synthesis, resulting in a high rate of mutations.